Anonymous Peer Benchmarks

Peer benchmarks show how your governance evidence compares to other organizations like you — without either side seeing the other's data. Comparisons are computed from sealed evidence, grouped into anonymous cohorts, and published only once enough firms contribute.

What you get

Each metric shows your current value, your percentile inside your cohort, and the cohort's quartile spread (25th, 50th, 75th and 90th percentile). Nothing identifies another organization — you never see peer names, counts per firm, or raw peer records.

Evidence fingerprint coverage — how much of your evidence is cryptographically fingerprinted.
Sealed-day coverage over the last 90 days — how consistently your evidence history is sealed.
Control review currency — the share of controls reviewed within their review window.
Nonconformity closure rate — how reliably findings get closed.
Mean incident resolution time — how quickly AI incidents are resolved.
Lineage depth per control — how well controls are connected to systems, use cases and datasets.

How cohorts are formed

A cohort is a grouping of organizations that are comparable to you. Cohort tags are assigned by Rhindon Cyber rather than self-declared, so a firm cannot pick a flattering peer group.

Industry sector: Your NAICS sector, chosen from the standard 20-sector taxonomy used across RAIC.
Size band: An employee-count band (for example 51–250 or 1,001–5,000).
Region: North America, EMEA, APAC, LATAM, or Global.
Framework scope: The frameworks you actively govern against — ISO 42001, the EU AI Act, NIST AI RMF, NIST CSF, SOC 2, CIS, HIPAA, or FINRA — so you are compared to firms doing the same work.

Why a comparison may be hidden

A cohort is only published once a sufficient number of contributing firms are in it. Below that floor the comparison is suppressed entirely, because a small cohort could let one organization infer another's posture. If you see "no peer comparison is available yet", your cohort has not reached that threshold — it will appear automatically when it does.

Suppression happens server-side; suppressed cohorts are never sent to the browser.
Your own data still contributes to other firms' cohorts in aggregate form only.
Cohort membership is re-evaluated as tags and contributor counts change.

How to read a percentile

75th percentile or above — you are ahead of most comparable firms on that metric.
40th to 74th — you are within the normal range for your cohort.
Below the 40th — the metric is a likely first target for improvement.
For time-based metrics such as incident resolution, lower values are better and the percentile is oriented accordingly.

Where benchmarks appear

AI Maturity dashboard — alongside your maturity index scores.
Registers and the lineage console — as peer context next to your own figures.
Board and executive reports — as an optional comparison section.

Availability

Peer benchmarks are included with Professional and Enterprise plans. On Starter, the benchmark card shows an upgrade prompt instead of cohort figures.